一种基于K-means改进蚁群算法的船舶航线设计方法

Method for designing ship routes based on improved ant colony algorithm of K-means

文章提出了一种基于K-means改进蚁群算法的船舶航线设计方法。首先利用K-means算法对栅格化海图情况进行聚类,模拟真实海况并得到多个具有区分特性标识的栅格类;然后将每一个类分别看成一个独立的TSP问题(Traveling Salesman Problem,旅行商问题),将蚁群算法应用在每个类内部和类之间,从而规划出一条船舶航线最短且能够安全避障的优化航线。通过仿真实验结果对比表明,该算法能够有效完成航线规划任务,并具有精度高、计算速度快等特点。

This paper proposes a ship route design method based on the combination of K-means clustering and ant colony algorithm.First use the K-means algorithm to cluster rasterized chart situations,simulate real sea conditions and obtain multiple grid classes with distinguishing characteristics;then treat each class as an independent TSP problem(Traveling Salesman Problem),the ant colony algorithm is applied within each class and between classes,so as to plan an optimized route with the shortest shipping route and safe avoidance of obstacles.The comparison of simulation results shows that the algorithm can effectively complete the route planning task,and has the characteristics of high accuracy and fast calculation speed.